ISSN 2250-0987

Kumar et al, UNIASCIT, Vol 2 (3), 2012, 292-300

LOAD FLOW BASED RELIABILITY

ASSESSMENT OF A REAL TIME RADIAL
DISTRIBUTION SYSTEM - CASE STUDY
N.M.G Kumar1, P Ramanjaneyulu Reddy 3, Dr.P.Sangameswara Raju2
1
Research scholar, Department of EEE, S.V.U. College of Engineering.&
Associate Professor, Department of EEE., Sree Vidyanikethan Engineering College.
3
PG student, Department of EEE., Sree Vidyanikethan Engineering College
Tirupati, Andhra Pradesh, India.
2
Professor, Department of E.E.E., S.V.U. College of Engineering,

Abstract In this paper mainly describe about the
Distribution system reliability is evaluated in two ways. One
by placing capacitor at weak voltage nodes for improvement
of voltage profiles, reducing the total losses. Second way by
improving reliability indices by placing protective equipment
(isolators)in the feeder. This paper present an effective
approach for real time evaluation of distribution power flow
solutions with an objective of determining the voltage profiles
and total losses, and to improve the voltage profiles and
reducing losses by placing capacitors at weak voltage profile
nodes using Particle Swarm Optimization (PSO) technique.
The Distribution System Reliability Indices are also
calculated for the existing radial distribution system before
and after placement of isolator. The paper presents a
topological characteristic of distribution networks have been
fully utilized to make the direct load flow solution is possible
for a real time system. Two matrices the bus-injection to
branch-current matrix and the branch-current to bus voltage
matrix and a simple matrix multiplication are used to obtain
power flow solutions. This paper also presents an approach
that determines optimal location and size of capacitors on
existing radial distribution systems to improve the voltage
profiles and reduce the active power loss. Capacitor sizing
was done by using Particle Swarm Optimization. In this
paper we have considered the load diversity factor for
analysis of load data for real time system. The performance
of the method was investigated on a real time distribution
system as case study. For load data of real time distribution
system the average power factor for feeder was considered
and distributed depending on the connected load. The load
flow results obtained are compared with power summation
method.
Keywords BIBC, BCBV, Diversity Factor, Reliability
Indices, Distribution Load Flows, PSO.
I. .INTRODUCTION
The demand for electrical energy is ever increasing. Today
over 21% (theft apart!!) of the total electrical energy
generated in India is lost in Transmission (5-7%) and
Distribution (15-18%). The electrical power deficit in the
country is currently about 35%. Clearly, reduction in
losses can reduces this deficit significantly. It is possible to
bring
down the distribution losses to 6-8% level in India with the
help of newer technological options (including information
technology) in the Electrical Power Distribution Sector
which will enable better monitoring and control. The
electric utility system is usually divided into three
subsystems which are Generation, Transmission, and
Distribution. A fourth division, which sometimes made is
Sub-Transmission. However, the latter can really be
considered as a subset of transmission since the voltage
levels and protection practices are quite similar. Electricity
distribution is the final stage in the delivery of electricity
to end users. A Distribution Network carries electricity
from the transmission system and delivers it to consumers.
Typically, the network would include medium-voltage
(less than 50 kV) power lines, electrical substations and
pole-mounted transformers, low-voltage (less than 1000
V) distribution wiring and sometimes electricity meters.
Electric power is normally generated at 11-25kV in a
power station. To transmit over long distances, it is then
stepped-up to 400kV, 220kV or 132kV as necessary.
Power is carried through a transmission network of high
voltage lines. Usually, these lines run into hundreds of
kilometers and deliver the power into a common power
pool called the grid. The grid is connected to load centers
through a sub-transmission network of normally 33kV (or
sometimes 66kV) lines. These lines terminate into a 33kV
(or 66kV) substation, where the voltage is stepped-down to
11kV for power distribution to load points through a
distribution network of lines at 11kV and lower.
The power network, which generally concerns the
common man is the distribution network of 11kV lines or
feeders downstream of the 33kV substation. Each 11kV
feeder which emanates from the 33kV substation branches
further into several subsidiary 11kV feeders to carry power
close to the load points (localities, industrial areas,
villages, etc.,). At these load points, a transformer further
reduces the voltage from 11kV to 415V to provide the last-
mile connection through 415V feeders (Low Tension (LT)
feeders) to individual customers, either at 240V (as single-
phase supply) or at 415V (as three-phase supply). A
feeder could be either an overhead line or an underground
cable. In urban areas, owing to the density of customers,
the length of an 11kV feeder is generally up to 3 km. On
the other hand, in rural areas, the feeder length is much
larger (up to 20 km). A 415V feeder should normally be
restricted to about 0.5 - 1.0 km. unduly long feeders lead
to low voltage at the consumer end.
II. DIVERSITY FACTOR
The probability that a particular piece of equipment will
come on at the time of the facility's peak load. It is
the ratio of the sum of the individual non-coincident

292

maximum demands of various subdivisions of the system
to the maximum demand of the complete system. The
diversity factor is always greater than 1. The (unofficial)
term diversity, as distinguished from diversity factor refers
to the percent of time available that a machine, piece of
equipment, or facility has its maximum or nominal load or
demand (a 70% diversity means that the device in question
operates at its nominal or maximum load level 70% of the
time that it is connected and turned on). This diversity
factor is used to estimate the load of a particular node in
the system.
III. LOAD FLOW STUDIES
The load-flow study in a power system has great
importance because it is the only system which shows the
electrical performance and power flow of the system
operating under steady state. Load-Flow studies are
performed to determine the steady-state operation of an
electric power system. A load-flow study calculates the
voltage drop on each feeder, the voltage at each bus, and
the power flow in all branch and feeder circuits. Losses in
each branch and total system power losses are also
calculated. Load-Flow studies are used to determine the
remain within specified limits, under various contingency
conditions only.
Load-flow studies are often used to identify the need for
additional Generation, Capacitive/Inductive VAR support
or the placement of capacitors and/or reactors to maintain
system voltages within specified limits. An efficient load-
flow study plays vital role during planning of the system
and also for the stability analysis of the system.
Distribution networks have high R/X ratio whereas the
transmission networks have high X/R ratio. Hence, the
distribution networks are ill-conditioned in nature.
Therefore, the variables for the load-flow analysis of
distribution systems are different from those of
transmission systems. The Transmission load flows
Newton Raphson Method; Gauss method fails due to high
X/R ratio. Many modified versions of the conventional
load-flow methods have been suggested for solving power
networks with high R/X ratio. The following are the
effective load flow techniques used in the distribution
networks:
1. Single-Line Equivalent Method
2. Very Fast Decoupled Method
3. Ladder Technique
4. Power Summation Method
5. Backward and Forward Sweeping Method
The proposed algorithm is tested on a Real Time system.
The remaining part of the paper is organized as follows:
Section 4 gives the formulation of load ow model,
Section 5 discusses the load flow algorithm, Section
5develops the test system and Real Time Systems results
and discussions and Section 6discusses the conclusions.
IV. FORMULATION OF LOAD FLOW
MODEL
(a) Algorithm Development:
The technique is based on two derived matrices, the bus-
injection to branch-current matrix and the branch current
to bus-voltage matrix, and equivalent current injections. In
ISSN 2250-0987
Kumar et al, UNIASCIT, Vol 2 (3), 2012, 292-300
this section, the development procedure will be described
in detail. For distribution networks, the equivalent current-
injection based model is more practical [5-13]. For bus ,
the complex load S is expressed by
Si=Pi+jQi ------------ (1)
Where I = 1, 2, ... N
And the corresponding equivalent current injection at the
kth iteration of solution is
Iik= (Pi+jQi/Vik)* --------- (2)
Where Vik and Iikare the bus voltages and equivalent
current injection of bus i at kth iteration respectively.
(b) Relationship Matrix Development
Fig.1. Simple distribution system
A simple distribution network shown in fig.1 is used as an
example the current equations are obtained from the
equation (2) .The relationship between bus currents and
branch currents can be obtained by applying Kirchhoffs
current law (KCL) to the distribution network. Using the
algorithm of finding the nodes beyond all branches
proposed by Gosh et al. The branch currents then are
formulated as functions of equivalent current injections for
example branch currents B1, B3 and B5 can be expressed
as
B1= I2+I3+I4+I5+I6
B3=I4+I5 ------------ (3)
B5= I6
Therefore the relationship between the bus current
injections and branch currents can be expressed as
= ------
(4a)
Eq (4a) can be expressed in general form as
[B]= [BIBC] [I] --------------------- (4b)
Where BIBC is a bus injection to branch current matrix,
the BIBC matrix is a upper triangular matrix and contains
values of 0 and 1 only.
The constant BIBC matrix is an upper triangular matrix
and contains values of 0 and 1 only.
The relationship between branch currents and bus voltages
as shown in Fig. 1. For example, the voltages of bus 2, 3,
and 4 are
V2=V1-B1Z12 ------------ (5a)
V3=V2-B2Z23 ------------ (5b)
V4=V3-B3Z34 ------------- (5c)
where Vi is the voltage of bus i, and Zij is the line
impedance between bus i and bus j. Substituting (5a) and
(5b) into (5c), (5c) can be rewritten as
V4=V1-B1Z12-B2Z23-B3Z34 ----------- (6)
293

From (6), it can be seen that the bus voltage can be
expressed as a function of branch currents, line
parameters, and the substation voltage. Similar procedures
can be performed on other buses; therefore, the
relationship between branch currents and bus voltages can
be expressed as
- = ---
(7a)
Equation can be rewritten in general form as:
[V]= [BCBV] [B] ------------- (7b)
Where BCBV is the branch current to bus voltage
matrix.
(C) Building Formulation Development:
Observing (4), a building algorithm for BBIBC matrix can
be developed as follows:
Step 1) For a distribution system with m-branch section
and n bus, The dimension of the BIBC matrix is m (n-1).
Step2) If a line section (B) is located between bus i and
bus j, copy the column of the Ith bus of the BIBC matrix to
the column of the jth bus and fill a 1 to the position of the
kth row and the jth bus column.
Step 3) Repeat Step (2) until all line sections is included in
the BIBC matrix. From (7), a building algorithm for
BCBV matrix can be developed as follows.
Step 4) For a distribution system with m-branch section
and n-k bus, the dimension of the BCBV matrix is m x (n-
1).
Step 5) If a line section is located between bus i and bus j
,copy the row of the ith bus of the BCBV matrix to the row
of the jth bus and fill the line impedance (Z ) to the position
of the jth bus row and the kth column.
Step 6) Repeat procedure (5) until all line sections is
included in the BCBV matrix. The algorithm can easily be
expanded to a multiphase line section or bus.
For example, if the line section between bus and bus is a
three-phase line section, the corresponding branch current
B will be a 31 vector and the in the BIBC matrix will be
a 33 identity matrix. Similarly, if the line section between
bus i and bus j is a three-phase line section, the ZI in the
BCBV matrix is a 33 impedance matrix.It can also be
seen that the building algorithms of the BIBC and BCBV
matrices are similar. In fact, these two matrices were built
in the same subroutine of our test program. Therefore, the
computation resources needed can be saved. In addition,
the building algorithms are developed based on the
traditional bus-branch oriented database; thus, the data
preparation time can be reduced and the proposed method
can be easily integrated into the existent DA.
D. Solution Technique Developments
The BIBC and BCBV matrices are developed based on the
topological structure of distribution systems. The BIBC
matrix represents the relationship between bus current
injections and branch currents. The corresponding
ISSN 2250-0987
Kumar et al, UNIASCIT, Vol 2 (3), 2012, 292-300
variations at branch currents, generated by the variations at
bus current injections, can be calculated directly by the
BIBC matrix. The BCBV matrix represents the
relationship between branch currents and bus voltages.
The corresponding variations at bus voltages, generated by
the variations at branch currents, can be calculated directly
by the BCBV matrix. Combining (4b) and (7a), the
relationship between bus current injections and bus
voltages can be expressed as
[V]=[BCBV][BIBC][I]=[DLF][I] ------(8)
And the solution for distribution power flow can be
obtained by solving (12) iteratively
Iik=Iir(Vik)+jIii(Vik)=((Pi+jQi)/Vik)* --------------(9a)
[Vk+1]=[DLF][Ik] -------------(9b)
[Vk+1] = [V] + [Vk+1] ----------- (9c)
According to the research, the arithmetic operation number
of LU factorization is approximately proportional to N 3.
For a large value of N, the LU factorization will occupy a
large portion of the computational time. Therefore, if the
LU factorization can be avoided, the power flow method
can save tremendous computational resource. From the
solution techniques described before, the LU
decomposition and forward/backward substitution of the
Jacobian matrix or the Y admittance matrix are no longer
necessary for the proposed method. Only the DLF matrix
is necessary in solving power flow problem. Therefore, the
proposed method can save considerable computation
resources and this feature makes the proposed method
suitable for online operation.
D. Losses Calculation
The Real power loss of the line section connecting
between buses i and i+1is computed as
P 2i Q 2i ------------ (10)
P RLOSS(i, i 1) R i ,i 1
|| Vi ||2
The Reactive power loss of the line section connecting
between buses i and i+1is computed as
P 2i Q 2i ----------- (11)
PXLOSS (i, i 1) X i ,i 1
|| Vi ||2
The total Real and Reactive power loss of the feeder
PFRLOSS
is determined by summing up the losses of all sections of
the feeder, which is given by:
N 1
PFRLOSS (i, i 1) PR LOSS (i, i 1) --------- (12)
i 1
N 1
PFXLOSS (i, i 1) PX LOSS (i, i 1) -------- (13)
i 1
V. PARTICLE SWARM OPTIMIZATION &
RELIABILITY INDICES
Particle Swarm Optimization (PSO) is a Meta heuristic
parallel search technique used for optimization of
continues nonlinear problems. The method was discovered
through simulation of a simplified social model. PSO has
roots in two main component methodologies perhaps more
obvious are ties to artificial life in general, and to bird
flocking, fish schooling and swarming theory in
294

particular.It is also related, however to evolutionary
computation and has ties to both genetic algorithms and
evolutionary programming. It requires only primitive
mathematical operators, and is computationally
inexpensive in terms of both memory requirements and
speed.
It conducts searches using a population of particles,
corresponding to individuals. Each particle represents a
Candidate solution to the capacitor sizing problem. In a
PSO system, particles change their positions by flying
around a multi-dimensional search space until a relatively
unchanged position has been encountered, or until
computational limits are exceeded. In social science
context, a PSO system combines a social and cognition
models. The general elements of the PSO are briefly
explained as follows:
Particle X(t): It is a k-dimensional real valued vector
which represents the candidate solution. For an ith particle
at a time t, the particle is described as Xi(t)={Xi,1(t),
Xi,2(t),... Xi,k(t)}.
Population: It is a set of n number of particles at a time t
described as {X1(t), X2(t) Xn(t)}.
Swarm: It is an apparently disorganized population of
moving particles that tend to cluster together while each
particle seems to be moving in random direction.
Particle Velocity V(t): It is the velocity of the moving
particle represented by a k-dimensional real valued vector
Vi(t)= {vi,1(t), vi,2(t) vi,k(t)}.
Inertia weight W(t): It is a control parameter that is used
to control the impact of the previous velocity on the
current velocity.
Particle Best (pbest): Conceptually pbest resembles
autobiographical memory, as each particle remembers its
own experience. When a particle moves through the search
space, it compares its fitness value at the current position
to the best
value it has ever attained at any time up to the current
time. The best position that is associated with the best
fitness arrived so far is termed as individual best or
Particle best. For each Particle in the swarm its pbest can
be determined and updated during the search.
Global Best (gbest): It is the best position among all the
individual pbest of the particles achieved so far.
Velocity Updation: Using the global best and individual
best, the ith particle velocity in kth dimension is updated
according to the following equation.
V[i][j]=K*(w*v[i][j]+c1*rand1*(pbestX[i][j]-
X[i][j])+c2*rand2*(gbestX[j]-X[i][j])).
where, K constriction factor, c1, c2 weight factors
w Inertia weight parameter , i particle number
j control variable ,
rand1, rand2 random numbers between 0 and 1
Stopping criteria: This is the condition to terminate the
search process. It can be achieved either of the two
following methods:
i. The number of the iterations since the last change of the
best solution is greater than a pre-specified number.
ii. The number of iterations reaches a prespecified
maximum value.
ISSN 2250-0987
Kumar et al, UNIASCIT, Vol 2 (3), 2012, 292-300
VI. ALGORITHM FOR PARTICLE SWARM
OPTIMIZATION:
Step1: Run the base case distribution load flow and
determine the active power loss.
Step2: Identify the candidate buses for placement
capacitor.
Step 3: Generate randomly n number of particles where
each particle is represented as
particle[i] ={Qc1,Qc2,Qc3,Qcj}
Step 4: Run the load flow by placing a particle i at the
candidate bus for reactive power compensation and store
the active power loss(TLP).
Step 5: Evaluate the fitness value. If the current fitness
value is greater than the its pbest value, then assign the
pbest value to the current value.
Step6:Determine the current global best (g_best_particles)
minimum among the particles individual best(pbest)
values.
Step 7: Compare the global position with previous. If the
current position is greater than the previous, then set the
global position to the current global position.
Step 8: update the particle velocity by using
V[i][j]=K*(w*v[i][j]+c1*rand1*(pbestX[i][j]-
X[i][j])+c2*rand2*(gbestX[j]-X[i][j])).
Step 9: Update the position of particle by adding the
velocity v[i][j].
Step 10: Now run the load flow and determine the active
power loss(pl) with the updated particle.
Step 11: Repeat step 5 to 7
Step 12: Repeat the same procedure for each particle from
step 4 to step 7.
VII. RELIABILITY INDICES:
System Average Interruption Duration Index (SAIDI)
The most often used performance measurement for a
sustained interruption is the System Average Interruption
Duration Index (SAIDI). This index measures the total
duration of an interruption for the average customer during
a given period. SAIDI is normally calculated on either
monthly or yearly basis; however, it can also be calculated
daily, or for any other period.
To calculate SAIDI, each interruption during the time
period is multiplied by the duration of the interruption to
find the customer-minutes of interruption. The customer-
minutes of all interruptions are then summed to determine
the total customer-minutes. To find the SAIDI value, the
customer-minutes are divided by the total customers. The
formula is,
Sumofcustomer int erruptionduration
U i * Ni
(14)
SAIDI
N
Tota ln umberofcustomers
i
Where
Ui=Annual outage time, Minutes
Ni=Total Number of customers of load point i.
SAIDI is measured in units of time, often minutes or
hours. It is usually measured over the course of a year, and
295

according to IEEE Standard 1366-1998 the median value
for North American utilities is approximately 1.50 hours.
Customer Average Interruption Duration Index
(CAIDI)
Once an outage occurs the average time to restore service
is found from the Customer Average Interruption Duration
Index (CAIDI). CAIDI is calculated similar to SAIDI
except that the denominator is the number of customers
interrupted versus the total number of utility customers.
CAIDI is,
U * Ni
i
Sumofcustomer int erruptionduration
SAIDI
* N
Tota ln umberofcustomer int erruptions
i i
Where
Ui=Annual outage time, Minutes
Ni= Total Number of customers of load point i.
i=Failure Rate.
CAIDI is measured in units of time, often minutes or
hours. It is usually measured over the course of a year, and
according to IEEE Standard 1366-1998 the median value
for North American utilities is approximately 1.36 hours
System Average Interruption Frequency Index (SAIFI)
The System Average Interruption Frequency Index
(SAIFI) is the average number of time that a system
customer experiences an outage during the year (or time
period under study). The SAIFI is found by divided the
total number of customers interrupted by the total number
of customers served. SAIFI, which is dimensionless
number, is
Tota ln umberofcustomer int erruptions * N i
SAIDI
N
Tota ln umberofcustomersserved
SAIFI
SAIDI
CAIDI
Where
Ni=Total Number of customers interrupted.
i=Failure Rate.
SAIFI is measured in units of interruptions per customer.
It is usually measured over the course of a year, and
according to IEEE Standard 1366-1998 the median value
for North American utilities is approximately 1.10
interruptions per customer.
Average Service Availability Index (ASAI)
The Average Service Availability Index (ASAI) is the
ratio of the total number of customer hours that service
was available during a given period of the total customer
hours demanded. This is sometimes called the service
reliability index. The ASAI is usually calculated on either
a monthly basis (730 hours) or a yearly basis (8,760
hours), but can be calculated for any time period. The
ASAI is found as,
(r * N )
i i
ASAI [1 ( )]*100
( NT * T )
ASUI 1 ASAI . (19)
Where
T= Time period under study, hours.
ri=Restoration Time, Minutes
Ni=Total Number of customers interrupted.
NT=Total Customers served.
ISSN 2250-0987
Kumar et al, UNIASCIT, Vol 2 (3), 2012, 292-300
Average Energy Not Supplied (AENS)
This is also called as Average System Curtailment Index
(ASCI)
Totalenergynot sup plied L *U (i)
a (i )
AENS . (20)
N
Tota ln umberofcustomersserved
i
VIII. INVESTIGATED REAL TIME SYSTEM
&RESULTS
The real time system considered in this paper is an
Eramitta (urban feeder) feeder located at mangalam
substation in Tirupati,Chittoor(Dt),Andhra
(15)
Pradesh,India.This feeder is not installed by any capacitor
banks. It is an fast growing residential area. In future if the
demand increases in this feeder then this feeder must be
installed by capacitor banks by using PSO technique.
Real time radial feeder system data
The radial distribution systems have following
characteristics
Base Voltage = 11KV. Base MVA=100.
Conductor type = All Aluminum Alloy Conductor
(AAAC)
Resistance = 0.55 ohm/KM.,
Reactance = 0.351 ohm/KM.
MATLAB program was developed for Load flow, PSO for
placement of capacitor to analyzing the results for Radial
Distribution feeder. To demonstrate the effectives of the
proposed concept, a 20-node 11kv Eramitta urban
distribution feeder is selected. Line data for this feeder is
shown in Table I.
i Load is not constant throughout the day; it varies from
(16) time to time. By considering the terms Diversity factor and
i
Power Factor five deferent conditions are consered.
. (17)
1.Average DF Good PF. 2.High DF High PF.
3.High DF Low PF. 4.Low DF High PF.
5.Low DF low PF
General condition that occurs is Average DF Good PF
where Average DF is 0.60 and Good PF is 0.92. When the
load is high (High DF) and the PF is also high (High PF),
this condition does not occur in the day but for the analysis
only it considered. When the load is high (High DF), the
PF decreases (Low PF), this condition occurs during the
peak demand.
When the load in Low (Low DF) then the PF is high (High
PF), this condition occurs during the light load conditions.
Low DF and Low PF condition does not occur in the day.
This condition is assumed for analysis only. The load flow
calculations are performed to get the voltages at each node
& the total power losses for 5 conditions. The voltage
magnitudes and the power losses are obtained by solving
. (18)
the simple algebraic equations which are illustrated in
section-III. Table II and Table III gives the voltage
magnitudes at each node and losses respectively for
Eramitta feeder at different conditions. The nodes which
are close to the source are having the higher voltage
magnitude is the nodes that are far away from the source
are of lower voltage magnitude (due to higher drop in
voltage).
296
 ISSN 2250-0987
Kumar et al, UNIASCIT, Vol 2 (3), 2012, 292-300

TABLE II
Voltage magnitudes at different conditions by BIBC & BCBV method
Bus Avg DF High F High DF Low DF Low DF
No gud PF High PF Low PF High PF Low PF
1 1 1 1 1 1
2 0.995 0.9918 0.9904 0.9953 0.9946
3 0.9909 0.9853 0.9827 0.9916 0.9902
4 0.988 0.9804 0.977 0.9888 0.987
5 0.9835 0.9734 0.9688 0.9848 0.9823
6 0.9833 0.973 0.9683 0.9846 0.9821
7 0.9824 0.9718 0.9669 0.9839 0.9812
8 0.982 0.9713 0.9662 0.9835 0.9807
9 0.9827 0.9719 0.9672 0.9838 0.9812
10 0.9807 0.969 0.9637 0.9821 0.9792
11 0.9806 0.9688 0.9635 0.982 0.9791
12 0.9804 0.9685 0.9632 0.9818 0.9789
13 0.9802 0.9681 0.9627 0.9815 0.9785
14 0.9807 0.969 0.9637 0.9821 0.9792
15 0.981 0.9692 0.9641 0.9822 0.9794
16 0.9804 0.9683 0.9632 0.9817 0.9789
17 0.9801 0.9675 0.9624 0.9813 0.9785
18 0.9821 0.9714 0.9664 0.9835 0.9808
19 0.9842 0.9745 0.9701 0.9854 0.983
20 0.9813 0.9698 0.9647 0.9826 0.9798

The radial diagram of 20-node Eramitta feeder shown in

Figure 3 is having 2-laterals to the main feeder. The first
lateral (5, 6, 7, 18, 8) and the second lateral (10, 12, 13)
are close to the source, so the voltage magnitudes at each
node are higher).This Eramitta feeder is not installed by
any capacitor bank at LT side. Without installation also
there are no nodes having voltages less than 0.95 p.u. So
there is no need of capacitor placement.
The improved voltage magnitudes by Single placement of
capacitor at 7th node and the voltage magnitudes by
multiple placement of capacitor using PSO at node 7 th and
8th are tabulated in Table VII.

Fig 2: Eramitta Radial feeder as per standard system

TABLE I TABLE III

Line data of Eramitta Feeder Losses at different conditions
Avg DF Good PF
Bus From To Distance R X
TLP = 27.4942 KW
No Node Node (KM) TLQ = 17.5463 KW
1 1 2 0.5 0.28 0.176 TL = 45.0405 KW
2 2 3 0.4 0.22 0.14 High DF High PF
3 3 4 0.3 0.17 0.105 TLP = 74.5279 KW
4 4 19 0.4 0.22 0.14 TLQ = 47.5624 KW
5 19 9 0.3 0.17 0.105 TL = 122.0903 KW
6 9 20 0.3 0.17 0.105 High DF Low PF
7 20 15 0.2 0.11 0.07 TLP = 100.5572 KW
8 15 16 0.5 0.28 0.176 TLQ = 64.1738 KW
TL = 164.731 KW
9 16 17 0.5 0.28 0.176
Low DF High PF
10 19 5 0.2 0.11 0.07
TLP = 24.7279 KW
11 5 6 0.1 0.06 0.035 TLQ = 15.7809 KW
12 6 7 0.4 0.22 0.14 TL = 40.5088 KW
13 7 18 0.2 0.11 0.07 Low DF Low PF
14 18 8 0.1 0.06 0.035 TLP = 32.7562 KW
15 20 10 0.2 0.11 0.07 TLQ = 20.9044 KW
16 10 11 0.2 0.11 0.07 TL = 53.6607 KW
17 10 14 0.2 0.11 0.07 In future if the demand increases (because it is an urban
18 10 12 0.2 0.11 0.07
feeder fast growing residential area) in the feeder then this
19 12 13 0.3 0.17 0.105
network must be installed by capacitor bank. For this

297
 ISSN 2250-0987
Kumar et al, UNIASCIT, Vol 2 (3), 2012, 292-300

condition i am assuming Unity DF with Low PF (Table
IV). The load flow results for Unity DF Low PF condition
are represented in Table V. From Table V, it can be found
that the following nodes are sensitive as voltages are less
than 0.95 p.u ,Nodes: 7,8,18,Voltages at sensitive nodes
can be improved by placing capacitor at single node or by
placing capacitor at multiple nodes. By using Particle
Swarm Optimization Technique (Section IV) the
capacitors are placed at multiple nodes and the voltage
profiles are shown in Figure 4. Compensated Nodes by
using PSO are tabulated in Table VI. By placing the 1.5
MVAR capacitor bank at 7th node the voltage profiles are
shown in Figure 4.
that when a capacitor placed at multiple nodes using PSO
the losses are much reduced when compared with single
capacitor placement at single node with highest capacity.
The details of the distribution system are shown in Table
IX. There are 4 interruption cases during the year 2011-
2012 (Table X). When the feeder was not provided with
isolators, 14 load points got affected during the 4
interruptions. The Distribution System Reliability Indices
are calculated by using section IV and are tabulated in
Table XI and the percentage of indices is represented in
pie chart as shown in Figure 5.
TABLE VI

TABLE IV Injected Reactive Power using PSO at different nodes

Load data (Unity DF and Low PF)
Unity DF Good PF Nodes Compensated 7,8
No of Total Connected
DTR DF=1 PF=0.866 Best Node=7 Best Particle -1106 KVAR
Customers Load
P(KW) Q(KVAR)
Best Node=8 Best Particle -377 KVAR
1 0 0 0 0
Total Injected Reactive Power -1483 KVAR
2 42 44.21 44.21 45.1
3 229 293.35 293.35 258.71
4 337 394.28 394.28 295.71
5 105 138.15 138.15 140.94 TABLE VII
6 105 130.72 130.72 98.04 Voltages for Single and Multiple Placement (PSO) of capacitor
7 213 283.9 283.9 289.64
8 533 603.25 603.25 615.44 Single Multiple
9 144 139.48 139.48 142.3 BEFORE
Placement Placement
10 162 185.71 185.71 189.46 Bus No Voltages Voltages Voltages
11 246 312.82 312.82 275.88 1 1 1 1
12 295 407.93 407.93 305.94 2 0.9844 0.9867 0.9886
13 2 4 4 1.31 3 0.972 0.9762 0.9796
14 489 487.47 487.47 192.66 4 0.9633 0.9689 0.9735
15 46 53 53 17.42 5 0.9569 0.9634 0.9689
16 450 426.29 426.29 206.46 6 0.9553 0.9618 0.9672
17 413 391.49 391.49 232.3 7 0.9491 0.9557 0.9612
18 92 121.67 121.67 75.4 8 0.948 0.9546 0.96
19 0 0 0 0 9 0.9606 0.9662 0.9708
20 0 0 0 0 10 0.9601 0.9657 0.9703
11 0.9581 0.9646 0.97
12 0.9583 0.9648 0.9703
TABLE V 13 0.956 0.9639 0.9705
Voltage magnitudes for Unity DF Low PF condition 14 0.959 0.9646 0.9702
15 0.9597 0.9653 0.9701
Bus No BIBC & BCBV Method 16 0.9548 0.9605 0.9653
1 1 17 0.9537 0.9593 0.9642
2 0.9844 18 0.9486 0.9552 0.9606
19 0.9622 0.9677 0.9724
3 0.972
20 0.9606 0.9662 0.9708
4 0.9633
5 0.9569 TABLE VIII
6 0.9553
7 0.9491
Status of Networks Before and After Compensation
8 0.948 Single Placement of capacitor Multiple Placement of Capacitor
9 0.9606
Q _ LOSS = 152.8818 KVAR Q _ LOSS = 152.8818 KVAR
10 0.9601
11 0.9581 P _ LOSS = 239.5584 KW P _ LOSS = 239.5584 KW
12 0.9583 MIN_V=0.9480 MIN_V=0.9480
Rank= 7,8,18 Rank= 7,8,18
13 0.956
After Compensation
14 0.959
15 0.9597 Q _ LOSS = 115.9827 KVAR Q _ LOSS =108.8445 KVAR
16 0.9548 P _ LOSS = 181.7393 KW P _ LOSS 170.5541 KW
17 0.9537 MIN_V=0.9546 MIN_V=0.9600
18 0.9486 Rank=0 Rank=0
19 0.9622 When the feeder is provided with isolator at 20 th node, the
20 0.9606 load point 20 will only be affected and the number of load
The status of networks before and after compensation for points affected are reduced from 14 to 9 during 4
single placement of 1.5MVAR capacitor at node 7 and interruption cases. Distribution Reliability Indices are
multiple placement of capacitor using PSO at nodes 7 and shown in Table III. The percentage of indices is
8 are shown in Table VII. From Table VIII, it is observed represented in pie chart as shown in Figure 6.

298
 ISSN 2250-0987
Kumar et al, UNIASCIT, Vol 2 (3), 2012, 292-300
TABLE XI
Distribution system Reliability Indices without Isolator
SAIFI 0.817 interruptions/customer
SAIDI 7.726 hrs/customer
CAIDI 9.456 hrs/customer interruption
ASAI 0.999118
ASUI 0.000882
AENS 8.641 KWh/customer

Fig 4. Voltage Graphs for Before, Single placement and multiple

placement using PSO
Interruption data

Table IX
Details of Distribution System Fig 5. Percentage of Indices representation in pie chart
Average
Load No of Total Connected
Connected Table XII:
Points Customers Load(KW)
load(KW) Interruption effect in a calendar year (with isolator)
1 0 0 0 Interruption Load Point Duration
Cause of Interruption
2 42 44.21 1.0526 Case Affected (hrs)
3 229 293.35 1.281 1 11 3 DTR failure and
4 337 394.28 1.17 7 10 for replacement
5 105 138.15 1.3157 2 8 10
Newly erected DTR
6 105 130.72 1.2449 18 10
7 213 283.9 1.3329 3 13 3 DTR failure and
8 533 603.25 1.1318 20 10 for replacement
9 144 139.48 0.9686 15 10 Shifting of DTR from
4
10 162 185.71 1.1464 17 10 one place to another place
16 10 and shifting of customers
11 246 312.82 1.2716
12 295 407.93 1.3828 Table XIII
13 2 4 2 Distribution system Reliability Indices (with isolator)
14 489 487.47 0.9969 SAIFI 0.511 interruptions/customer
15 46 53 1.1522 SAIDI 4.667 hrs/customer
16 450 426.29 0.9473 CAIDI 9.130 hrs/customer interruption
17 413 391.49 0.9479 ASAI 0.999467
18 92 121.67 1.3225 ASUI 0.000533
AENS 5.059 KWh/customer
19 0 0 0
20 0 0 0
Total 3903 4417.71

TABLE X
Interruption effect in a calendar year (without isolator)
Interruption Load Point Duration
Cause of Interruption
Case Affected (hrs)
1 11 3 DTR failure and
7 10 for replacement
2 8 10
Newly erected DTR
18 10
3 13 3
DTR failure and
20 10
for replacement
10,14 10
11,12 10 Shifting of DTR from
4
13,15 10 one place to another place
16 10 and shifting
17 10 of customers Fig 6: Percentage of Indices representation in pie chart
When the feeder is not provided with isolator the Average
Energy Not Supplied (AENS) is 8.641 KWh/Customer.
When the feeder is provided with isolator at 20 th node the

299
 ISSN 2250-0987
Kumar et al, UNIASCIT, Vol 2 (3), 2012, 292-300

Average Energy Not Supplied (AENS) is reduced to 5.059 [19] IEEE Standards, IEEE Guide for Electric Power
Distribution Reliability Indices, IEEE Power Engineering Society.
KWh/Customer.
[20] Robert J.Rusch and David L.Metz, Customer Oriented Reliability
VII. CONCLUSIONS Indices and Data Collection, Stanley Consultants, Muscatine,IA 52761.
[21] Roy Billinton and Ronald N.Allan, A Text Book on
Eramitta feeder is applied with load flow and the weak voltage Reliability Avaluation of Power Systems second Edition, Plenum
profiles are identified and those nodes are proposed for capacitor Press, New York and London.
placement using particle swarm optimization technique. The [22] S. Ghosh and D. Das, Method for load-flow solution of radial
distribution Networks, IEEE Proceedings on Generation, Transmission
voltages and losses before and after compensation are tabulated. & Distribution, Vol.146, No. 6, pp. 641-648, 1999
By placement of capacitor at single node with highest capacity
the voltage profiles get improved and losses get reduced, but by Authors detail:
distributing the same capacity and placing the capacitors at Mr. N.M.G.KUMAR Currently pursuing
multiple nodes using Particle Swarm Optimization technique the P.hD at SVU College of engineering at
voltage profiles are good as comparing with single placement and Tirupati, AP, India and Obtained his B.E in
the losses are much more reduced as compared with single Electrical and Electronics Engineering from
placement of capacitor. From Table 11 XI and Table XIII, it is
Bangalore University at S.M.V.I.T.S.,
concluded that by providing more isolators in the feeder we can
reduce the Average Energy Not Supplied (AENS). Bangalore. Obtained M.Tech (PSOC) at
S.V.U.college engineering, Tirupati. Area of
REFERENCES: interest are power system planning, power system
[1] K.Prakash, M.Sydulu, Partical Swarm Optimization Based Capacitor optimizations, power system reliability studies, Real time
Placement on Radial Distribution System, IEEE Power Engineering.
Society general meeting June 2007, pp. 1-5, 2007. application of power system and like non-linear controllers
[2] A.A.A. Esmin and G. Lambert-Torres, A Particle Swarm applications to power systems.
Optimization Applied to Loss Power Minimization, IEEE Transactions
on Power Systems, USA, Vol.20, No.2, pp. 859-866, 2005. P RAMANJANEYULU REDDY currently pursuing his
[3] S.Gosh and D.Das, Method for load-flow solution of radial
distribution networks, IEE Proc.- Gener. Transm.Distrib.Vol. 146, No. M.Tech from Sree Vidyanikethan engineering college,
6, November 1999. tirupati. Obtained his B.Tech in Electrical and Electronics
[4] N. Balijepalli, S. S. Venkata, and R. D. Christie, Modeling and Engineering from JNTU Anantapur ,Anantapur University
analysis of distribution reliability indices, IEEE Trans. Power Del., at Audisankara college of engineerin, Gudur, Nellore,AP.
vol.19, no. 4, pp. 19501955, Oct. 2004.
[5] J. H. Teng andW. M. Lin, Current-based power lowsolutions for Areas of interest are power system analysis, application of
distribution systems, in Proc. IEEE Int. Conf. Power Syst. Technol., FACTS devices in Transmission systems. Distribution
Beijing, China, 1994, pp. 414418. systems and automation, Artificial Intelligent systems
[6] T. S. Chen, M. S. Chen, T. Inoue, and E. A. Chebli, Three-phase applications to power systems.
cogenerator and transformer models for distribution system analysis,
IEEE Trans. Power Delivery, vol. 6, pp. 16711681.2, Oct. 1991.
[7] T.-H. Chen, M.-S. Chen, K.-J. Hwang, P. Kotas, and E. A. Chebli, Dr.P.SANGAMESWARA RAJU is presently working
Distribution system power flow analysisA rigid approach, IEEE as professor in S.V.U.college engineering, Tirupati.
Trans. Power Delivery, vol. 6, pp. 11461152, July 1991. Obtained his diploma and B.Tech in
[8] T. H. Chen and J. D. Chang, Open wye-open delta and open delta-
open delta transformer models for rigorous distribution system analysis, Electrical Engineering, M.Tech in power
in Proc. Inst. Elect. Eng., vol. 139, 1992, pp. 227234. system operation and control and PhD in
[9] K. A. Birt, J. J. Graffy, J. D. McDonald, and A. H. El-Abiad, Three S.V.University, tirupati. His areas of interest
phase load flow program, IEEE Trans. Power Apparat. Syst., vol. are power system operation, planning and
PAS95, pp. 5965, Jan./Feb. 1976.
[10] D. Shirmohammadi, H. W. Hong, A. Semlyen, and G. X. Luo, A application of fuzzy logic to power system,
compensation- based power flow method for weakly meshed distribution application of power system like non-linear controllers.
and transmission networks, IEEE Trans. Power Syst., vol. 3, pp. 753
762, May 1988.
[11] G. X. Luo and A. Semlyen, Efficient load flow for large weakly
meshed networks, IEEE Trans. Power Syst., vol. 5, pp. 13091316,
Nov. 1990.
[12] C. S. Cheng and D. Shirmohammadi, A three-phase power flow
method for real-time distribution system analysis, IEEE Trans. Power
Syst., vol. 10, pp. 671679, May 1995.
[13] R. D. Zimmerman and H. D. Chiang, Fast decoupled power flow
for unbalanced radial distribution systems, IEEE Trans. Power Syst.,
vol. 10, pp. 20452052, Nov. 1995.
[14] W. M. Kersting and L. Willis, Radial Distribution Test Systems,
IEEE Trans. Power Syst., vol. 6, IEEE Distribution PlanningWorking
Group Rep., Aug. 1991.
[15] M. E Baran and F. F. Wu, Optimal Sizing of Capacitors Placed on a
Radial Distribution System, IEEE Trans. Power Delivery, vol. no.1, pp.
1105-1117, Jan. 1989.
[16] M. E. Baran and F. F. Wu, Optimal Capacitor Placement on radial
distribution system, IEEE Trans. Power Delivery, vol. 4, no.1, pp.
725734, Jan. 1989.
[17] M. H. Haque, Capacitor placement in radial distribution systems for
loss reduction, IEE Proc-Gener, Transm, Distrib, vol, 146, No.5, Sep.
1999.
[18] R.Billinton and J. E. Billinton, Distribution system reliability
indices,
IEEE Trans. Power Del., vol. 4, no. 1, pp. 561586, Jan. 1989.

300